Conflict Intensity and Economic Structure: A Unified Empirical and Theoretical Framework

Details

Supervisors

Maystadt, Jean-François ; Parodi, Francesca

Faculty

Faculté des sciences économiques, sociales, politiques et de communication

Degree label

Master [120] en sciences économiques, orientation générale, à finalité approfondie

Abstract

enThis thesis studies how armed conflict affects local economic activity across different productive structures in Africa. It combines a dynamic theoretical model with a spatial empirical analysis to examine whether the consequences of conflict differ between labor-intensive agricultural environments and capital-intensive mining environments. In the model, households allocate time between production and banditry. Conflict can generate short-run appropriation gains when looting is sufficiently attractive, but it also disrupts production, destroys productive capacity, lowers the return to capital, and weakens investment incentives. The model therefore predicts that conflict may temporarily increase measured activity in lootable environments, while generating persistent losses in sectors more exposed to disruption and capital destruction. Under sufficiently strong feedback between violence and declining productive capacity, conflict may contribute to the emergence of a conflict trap. The empirical analysis constructs a PRIO-GRID cell-year panel for African countries over the period 1989--2024, combining georeferenced conflict events from the UCDP Georeferenced Event Dataset with harmonized night-time lights, agricultural statistics from HarvestStat Africa, and mine-level production data. Local economic activity is proxied using harmonized night-time lights. Using an event-study framework, the thesis traces the dynamic response of local luminosity around conflict onset and allows this response to differ across agricultural yield groups and mining production groups. The results suggest that the economic effects of conflict are strongly heterogeneous across production structures. In agricultural areas, especially in high-yield cells, conflict onset is associated with short-run increases in measured night-time lights, consistent with an appropriation or reallocation mechanism. However, these gains attenuate over time, indicating that they are temporary rather than development-enhancing. In mining areas, particularly within major-war countries, conflict is associated with immediate and persistent declines in night-time lights, consistent with the disruption and capital-destruction channels emphasized by the model. Overall, the findings support the interpretation of conflict as a dynamic constraint on local development: even when conflict temporarily raises measured activity, it undermines the conditions for sustained growth and may push local economies onto persistently weaker trajectories.